Abstract Measurable residual disease (MRD) monitoring by multiparameter flow cytometry (MFC) is well established in bone marrow (BM) samples for acute myeloid leukemia (AML), but its use in peripheral blood (PB) remains less developed. We adapted a semi‐automated and well validated MFC‐MRD assay, originally developed for BM, to PB aiming to evaluate its applicability and concordance with established methods. To transfer the MFC‐MRD assay from BM to PB, we incorporated population‐specific reference values derived from healthy and chemotherapy‐exposed non‐AML controls. MRD detection was based on a combined aberrant leukemia associated immunophenotype (aLAIP) and different from normal (DfN) approach. We compared MFC‐MRD detection in paired PB and BM samples and evaluated concordance with molecular MRD (Mol‐MRD) diagnostics. Additional analyses included assessment of mast cells as markers for BM hemodilution as well as time requirements for analyses. Reference values for aberrant populations in PB were generally lower than in BM, particularly considering immature markers. At follow‐up, MFC‐MRD positivity was less common in PB than in BM, leading to a concordance rate of 73.7%. Both quantitative and qualitative differences in aberrant antigen expression between PB and BM were observed. Most cases of MRD positivity were characterized by newly emerging aberrant features rather than persistent ones. Concordance between MFC‐MRD from PB and Mol‐MRD from BM was 64.6%, similar to that of MFC‐MRD from BM versus Mol‐MRD from BM (58.5%). Most discordant cases (MFC‐MRD pos /Mol‐MRD neg ) may reflect clonal hematopoiesis or limited spectrum of molecular panels. Mast cell quantification proved useful in identifying hemodiluted BM samples, which comprised ~15% of cases. Inter‐rater agreement for MRD detection was strong ( Kα > 0.8). The MFC‐MRD gating and evaluation was rapid (~2 min/sample in both PB and BM). The adapted MFC‐MRD assay is feasible and robust in PB, offering a fast and reproducible alternative to BM‐based MRD monitoring. However, the prognostic relevance of the proposed method needs to be validated in a prospective cohort of patients.
Schadt et al. (Mon,) studied this question.